Low PAPR OTFS Using Data Sequence Substitution
Abstract
We propose a peak-to-average power ratio (PAPR) reduction scheme for Orthogonal Time Frequency Space (OTFS) modulation based on delay-Doppler (DD)-domain data sequence substitution. The proposed scheme replaces high-PAPR DD-domain row sequences with mapped low-PAPR sequences and recovers the original sequences at the receiver through inverse substitution. The Hungarian algorithm is used as an offline assignment tool for pairing high- and low-PAPR sequence sets, while a <inline-formula> <tex-math notation="LaTeX">$k$ </tex-math></inline-formula>-dimensional (<inline-formula> <tex-math notation="LaTeX">$k$ </tex-math></inline-formula>-d) tree reduces the online sequence-search burden. To address scalability, we further introduce a partitioned threshold-based implementation for large Doppler dimensions, where an <inline-formula> <tex-math notation="LaTeX">$N=32$ </tex-math></inline-formula> frame is divided into length-8 Doppler partitions and the proposed substitution is applied within each partition before low-PAPR OTFS (LP-OTFS)-style interleaving. For <inline-formula> <tex-math notation="LaTeX">$N=32$ </tex-math></inline-formula>, the proposed partitioned TDSM with <inline-formula> <tex-math notation="LaTeX">$\mathrm {PAPR}_{\mathrm {Th}}=4$ </tex-math></inline-formula> reduces the PAPR by about 2.46 dB relative to plain LP-OTFS with QPSK and provides additional BER improvement in the low-IBO region. With 16QAM, the corresponding PAPR gain is about 1.05 dB, and the low-IBO BER improvement is also retained.